Direct Aqueous Mineral Carbonation of Waste Slate Using Ammonium Salt Solutions

The carbonation of asbestos-containing waste slate using a direct aqueous mineral carbonation method was evaluated. Leaching and carbonation tests were conducted on asbestos-containing waste slate using ammonium salt (CH3COONH4, NH4NO3, and NH4HSO4) solutions at various concentrations. The CH3COONH4 solution had the highest Ca-leaching efficiency (17%–35%) and the NH4HSO4 solution had the highest Mg-leaching efficiency (7%–24%) at various solid dosages and solvent concentrations. The CaCO3 content of the reacted materials based on thermogravimetric analysis (TGA) was approximately 10%–17% higher than that of the as-received material for the 1 M CH3COONH4 and the 1 M NH4HSO4 solutions. The carbonates were precipitated on the surface of chrysotile, which was contained in the waste slate reacted with CO2. These results imply that CO2 can be sequestered by a direct aqueous mineral carbonation using waste slate.


Introduction
CO2 mineral carbonation is a method to permanently sequester CO2 as a form of carbonate minerals with or without aqueous phase.For the CO2 mineral carbonation process with aqueous phase, CO2 is reacted with raw materials containing alkaline earth metals (mostly Ca and Mg) in aqueous solutions (i.e., direct aqueous mineral carbonation) or is reacted with alkaline earth metals leached from raw materials in aqueous solutions (i.e., indirect aqueous mineral carbonation) to form carbonate minerals [1,2].
The mineral carbonation to sequester CO2 has been extensively studied due to its following advantages: (1) non-requirement for any underground geological storage sites; (2) permanent CO2 sequestration without long-term monitoring; (3) potential immobilization of toxic elements contained in raw materials; and (4) beneficial use of produced carbonates.However, the major challenge of the mineral carbonation method is to enhance the leaching capacity of alkaline earth metals from raw materials, which is a main factor affecting rate and degree of mineral carbonation [3,4].The leaching processes of alkaline earth metals from raw materials are generally expensive due to the need for increasing temperature and pressure, acid or base solutions, and grinding raw materials.
Previous researches regarding aqueous mineral carbonation have mainly focused on enhancing the leaching efficiency of alkaline earth metals by pre-treating raw materials before carbonation processes [4,[5][6][7][8].Natural alkaline materials and alkaline industrial wastes have been extensively evaluated for the mineral carbonation, e.g., [8,[9][10][11][12][13][14][15][16][17][18][19].Natural alkaline materials are relatively abundant compared with alkaline industrial wastes but their use requires pretreatment to enhance the leaching of the alkaline earth metal due to their strong chemical stability.Even if industrial wastes have more limited availability, the alkaline earth metals can be relatively easily leached from the alkaline industrial byproducts because of their chemical instability [20].
Asbestos-containing waste slate is one of the potential alkaline industrial wastes to sequester CO2 due to its high content of Ca and Mg.Asbestos-containing waste slate is mostly comprised of cement, which is a Ca source, and chrysotile, which is an Mg source.Chrysotile, which is considered as a carcinogenic material, is the fibrous magnesium silicate mineral (Mg3(Si2O5)(OH)4) in the serpentine group.Asbestos-containing slate was used as a roofing material in South Korea during 1960 and 1970s due to its high insulating capacity, but subsequently has not been legally allowed for use due to the toxicity of chrysotile, which is a form of asbestos.Currently, the asbestos containing waste slate is mostly disposed of in government certified landfills without reuse in South Korea.
Due to the health risks such as asbestosis, lung cancer, and mesothelioma, a number of studies suggested safe disposal schemes of asbestos containing materials.Most treatment methods of asbestos-containing materials focused on morphological alteration of asbestos using chemical and thermal treatment.Chemical treatment methods of asbestos-containing material using oxalic acid [21], Na-oxalate and Na-acetate [22], sulfuric acid [23], and hydrogen peroxide [24] could alter significantly asbestos material.Thermal treatment methods of asbestos using microwave [25][26][27] and microwave air plasma [28] could also alter effectively asbestos to non-hazardous form.However, thermal treatment requires vast amount of energy.Gualtieri and Tartaglia [29] reported that temperature higher than 1000 °C is required to entire transformation of asbestos to non-hazardous silicate glass phase.
Yoshikawa et al. [27] also showed that asbestos was completely transformed to non-hazardous form at the temperature >1000 °C with microwave treatment.
The main objective of this study was to evaluate the feasibility of CO2 sequestration using waste slate by a direct aqueous mineral carbonation.Ammonium salt (CH3COONH4, NH4Cl, NH4NO3, and NH4HSO4) solutions were used as solvents for a direct aqueous mineral carbonation of asbestos-containing waste slate.The leaching and carbonation behaviors of asbestos-containing waste slate were investigated at room temperature and atmospheric pressure conditions.For the carbonation tests, CO2 was injected into the mixture of waste slate and ammonium salt solution.Morphological alteration of asbestos-containing waste slate after carbonation was also evaluated.

Materials
Waste roofing slate panels were collected from an abandoned house in South Korea.The waste slate was broken using a hand hammer to get particle size less than 0.425 mm.The slate particles were used in the leaching and carbonation tests.CH3COONH4, NH4NO3, and NH4HSO4 (Sigma-Aldrich Co., St. Louis, Mo, USA) solutions were used as solvents.Deionized (DI) water was used to prepare the ammonium salt solutions.

Leaching Tests
The leaching test was performed with various slate solid dosages ranging between 20 and 150 g/L in a 50 mL polypropylene copolymer centrifuge tube using an ammonium salt solution for four hours.1-10 g of oven-dried slate was added into a 50 mL polypropylene copolymer centrifuge tube with 1, 2, and 4 M solutions.The mixture was shaken at room temperature and atmospheric pressure using a water bath shaker and the pH and electrical conductivity (EC) were then measured.The mixture was centrifuged at 5000 rpm for 30 min (VS-550i, Vision Scientific CO., Daejeon, Korea) and was then filtered using a 0.2 µm filter (ADVANTEC®, Advantec MFS, Inc. Tokyo, Japan).All of the leaching tests were conducted a single time.

Carbonation Tests
For the carbonation test, the waste slate and 1 M CH3COONH4 solution or 1 M NH4HSO4 solution at various solid dosages ranged from 20 to 150 g/L were mixed thoroughly for four hours in a 500 mL Elrenmeyer flask (Dongsung Scientific, Busan, Korea).The initial pH of the mixtures of the waste slate and 1 M CH3COONH4 solution or 1 M NH4HSO4 solution were about 9.4 and 9.2, respectively.A 15 vol.% CO2 gas mixture with 85 vol.% N2 was then injected at a flow rate of 200 mL/min into the slurry of waste slate and 1 M CH3COONH4 solution or 1 M NH4HSO4 solution at various solid dosages in the 500 mL Erlenmeyer flask.The 15 vol.% CO2 gas mixture was chosen because the flue gas from coal fired power plants in South Korea generally contains 15 vol.% CO2.During the carbonation tests, the slurry was mixed using a magnetic stirrer at a stirring rate of approximately 300 rpm.The procedure of the carbonation test used in this study is similar to that used in Jo et al. [8].The carbonation test was performed at a room temperature and atmospheric pressure.The CO2 injection was stopped at a pH of around 7.5 for preventing CaCO3 dissolution [30].All of the carbonation tests were conducted a single time.

Chemical Analysis and Material Characterization
The cation concentrations of filtrates from the leaching and carbonation tests were analyzed by inductivity coupled plasma-atomic emission spectroscopy (ICP-AES, OPTIMA 3000XL, Perkin Elmer, Wellesley, MA, USA).The elemental composition of the as-received material was determined by using microwave digestion.0.5 g of the waste slate raw material with 10 mL of HNO3 was digested using microwave digestion for 10 min at 175 °C and the leachate was then analyzed by ICP-AES (Perkin Elmer, Wellesley, MA, USA).
An X-ray diffractometer (Xpert MPD, Phillips, Almelo, The Netherlands) and a field emission scanning electron microscope (FE-SEM: S-4300, Hitachi, Tokyo, Japan) equipped with energy dispersive X-ray spectroscopy (EDX: Ex-20, Horiba, Kyoto, Japan) were used to characterize the as-received and reacted materials.The thermogravimetric analysis (TGA) (SDTG-60H, Shimadzu Corp., Kyoto, Japan) was performed on the reactant and products to determine the calcium carbonate (CaCO3) content.

Raw Material
The as-received waste slate contained mainly Ca (25.2%).(Table 1).Table 1 shows selected elemental composition determined by using microwave digestion.The as-received waste slate mainly consists of calcite, chrysotile, and Ca-Mg-Al silicates (Figure 1).The CaCO3 content of the waste slate determined using the results of TGA [8], as shown in Figure 2 was approximately 29.8%.The CaCO3 may have been formed by natural carbonation because the slate had been exposed to atmosphere for a very long period of more than 20 years.It is well known that cement based materials exposed to atmosphere can absorb CO2 during the service life [31].

Leaching Behaviors
The Ca-and Mg-leaching efficiencies after leaching tests as a function of solid dosage (20, 50, 100, 150, and 200 g/L) and solvent concentration (1, 2, and 4 M) are shown in Figures 3 and 4. The leaching efficiency was determined by following equation: where MMe-leached is the mass (g) of the metal in the leachate obtained after the leaching test, Mt is the total mass (g) of the material used in the leaching test, CMeO is the metal content of the material determined by the total elemental analysis.For all solvent conditions, the metal leaching efficiencies were decreased when the solid dosage increased from 20 to 200 g/L (Figures 3 and 4).Among the 1 M ammonium salts solutions, the CH3COONH4 and NH4NO3 solutions had higher Ca-leaching efficiencies than did the NH4HSO4 solution, regardless of solid dosage.Jo et al. [8] reported that 1 M NH4NO3 and CH3COONH4 solutions could effectively dissolve Ca ions from waste cement.The NH4HSO4 solution may have had a lower Ca-leaching efficiency than the NH4NO3 and CH3COONH4 solutions due to the lower Ca selectivity or the precipitation of gypsum (CaSO4) during the leaching step (Figure 5).In contrast, the NH4HSO4 solution had the highest Mg-leaching efficiency among all the solvents.Wang and Maroto-Valer [6] reported that 1.4 M NH4HSO4 solution leached 100% of Mg and 98% of Fe, but only 13.6% of Si, from serpentine at 100 °C.However, the lower Mg-leaching efficiency (<24%) obtained in this study was probably due to the lower temperature and solvent concentration.Nevertheless, Mg ions were leached dominantly from waste slate in the NH4HSO4 solution.For the 20 g/L of the solid dosage, the CH3COONH4 and NH4NO3 solutions had relatively high Ca-leaching efficiencies (~37%) at high solvent concentrations (>2 M) (Figure 4).However, the CH3COONH4 and NH4NO3 solutions had a slightly lower Ca-leaching efficiency (34%) at low solvent concentration (1 M).The NH4HSO4 solution had the lowest Ca-leaching efficiency at all solvent concentrations.On the contrary, the NH4HSO4 solution had the highest Mg-leaching efficiency (18%-20%) at all solvent concentrations.The Mg-leaching efficiencies of both CH3COONH4 and NH4NO3 solutions were lower than that of the NH4HSO4 solution but increased with increasing the solvent concentration.The Mg-leaching efficiency of NH4HSO4 was slightly affected by the solvent concentration, suggesting that the CH3COONH4 and NH4HSO4 solutions were the most efficient solvents for Ca and Mg leaching from waste slate, respectively.
The XRD patterns on the unreacted and reacted waste samples are shown in Figure 5.For the 1 M CH3COONH4 solution, the chrysotile peak intensity was slightly decreased, regardless of the solid dosage (Figure 5).Even though Ca ions were dissolved into the leachate, the calcite peak intensity barely changed after the leaching tests, possibly indicating that Ca-silicates in the waste slate was a main Ca source in the slurry.The quartz peak identified after the leaching tests was further possible evidence for the dissolution of Ca-silicates.
For the 1 M NH4HSO4 solution, gypsum (CaSO4• 2H2O) was identified at all solid dosages, suggesting that gypsum was precipitated during the leaching process.The gypsum peak intensity was higher at the solid dosage of 50 g/L and decreased with increasing the solid dosage.The calcite peak intensity also barely changed after the leaching test using the 1 M NH4HSO4 solution.In addition, the chrysotile peak was still observed after the leaching tests.The Mg-leaching efficiency (~20%) also confirmed that chrysotile, which was an Mg source, was not fully decomposed after the leaching tests using the CH3COONH4, NH4NO3, and NH4HSO4 solutions.

Carbonation Behaviors
The carbonation tests were carried out on suspensions from the leaching test using 1 M CH3COONH4 and 1 M NH4HSO4 solutions.The 1 M CH3COONH4 and NH4HSO4 solutions were selected due to their high metal leaching efficiency and selectivity in the leaching test.The 1 M CH3COONH4 solution had a high Ca-leaching efficiency and Ca selectivity and the 1 M solution NH4HSO4 showed high Mg-leaching efficiency and Mg selectivity (Figures 3 and 4).
The cation concentrations of filtrates obtained from the leaching and carbonation tests are shown in Table 2.After the leaching tests, the pH of mixtures using 1 M CH3COONH4 and 1 M NH4HSO4 solutions ranged between 9.0 and 9.8, which was suitable for mineral carbonation.For the CH3COONH4 solution, the Ca concentration decreased significantly after carbonation test due to the precipitation of CaCO3.For the NH4HSO4 solution, however, the Ca concentration was not greatly changed after the carbonation test.In contrast, the Mg concentration increased after the carbonation test for both CH3COONH4 and NH4HSO4 solutions, probably because Mg ions were released from the solid phase during the CO2 injection.Dissolution of Mg ion is favorable at pH < 8.0 [32].The Mg ion might dissolve further during carbonation because CO2 injection was stopped when pH of the suspension reached about 7.5.The Si concentration decreased after the carbonation for both CH3COONH4 and NH4HSO4 solution, probably due to the precipitation of SiO2 from dissolved Ca-Mg-Al-Si-O complex during the CO2 injection.The carbonation efficiency was determined using the Ca concentration of the filtrates obtained from the leaching and carbonation tests using the following equation: where MCa-leached is the mass (g) of Ca in the filtrate obtained from the leaching test, Mcarbonation is the mass (g) of Ca in the filtrate obtained from the carbonation test, Mt is the total mass (g) of the waste slate used in the tests, and CCa is the calcium content (%) of the waste slate determined by total elemental analysis (Table 1).The carbonation efficiency is shown in Figure 6.The 1 M CH3COONH4 solution had a higher Ca carbonation efficiency than the 1 M NH4HSO4 solution, regardless of the solid dosage.The Ca carbonation efficiencies for both 1 M CH3COONH4 and 1 M NH4HSO4 solutions decreased as the solid dosage increased, which was comparable to the Ca-leaching efficiency.

Characteristics of Reacted Asbestos-Containing Waste Slate
The CaCO3 content of the reacted materials determined using results of TGA is shown in Figure 7.The CaCO3 content of the reacted material was approximately 10%-17% higher than that of the as-received material, except for the 1 M NH4HSO4 solution, at a solid dosage of 100 g/L.These results suggest that a maximum CO2 sequestration capacity of ~0.048 kg CO2/kg waste slate was obtained at a solid dosage of 20 g/L using the 1 M CH3COONH4 solution.The CaCO3 content of the reacted material was slightly decreased with increasing the solid dosage from 20 to 150 g/L, except the solid dosage of 100 g/L, with the 1 M NH4HSO4 solution.The CaCO3 content in the 1 M CH3COONH4 solution was approximately 10% higher than that in the 1 M NH4HSO4 solution.On the other hand, the carbonation efficiency in the 1 M CH3COONH4 solution based on Ca concentrations in the filtrate was approximately 20% higher than that in the 1 M NH4HSO4 solution.These results suggest that Ca ions obtained during the leaching process and Ca ions obtained by dissolving further from waste slate might be precipitated during the CO2 injection.After the carbonation tests using 1 M CH3COONH4 and 1 M NH4HSO4 solutions, the calcite peak intensity increased at all the solid dosages (Figure 8).For both 1 M CH3COONH4 and 1 M NH4HSO4 solutions, a chrysotile peak was still observed in the reacted materials after the carbonation tests.For the 1 M CH3COONH4 and 1 M NH4HSO4 solutions, almost similar XRD results were obtained, except for the formation of gypsum (CaSO4• 2H2O) for the 1 M NH4HSO4 solution (Figure 8).The images of SEM analysis on the as-received and reacted materials obtained after the carbonation tests using 1 M CH3COONH4 and 1 M NH4HSO4 solutions are shown in Figures 9 and 10 waste slate, as shown in Figure 9a, even though carbonates were formed by natural carbonation due to the long atmospheric exposure of the slate panels.In addition, the SEM images confirmed that the chrysotile originally contained in the waste slate was not fully decomposed after the carbonation tests.
Mg ions from chrysotile can be dissolved at high pressure (>3 MPa) and temperature (>600 °C) conditions or moderate temperature (>70 °C) with very strong acid solutions (e.g., hydrochloric acid, oxalic acid, and sulfuric acid), e.g., [33,34].However, cubic calcite was precipitated on the surface of needle-like chrysotile at all solid dosages for both 1 M CH3COONH4 and 1 M NH4HSO4 solutions after the carbonation tests (Figures 9 and 10).Health impact of asbestos is mainly caused by its needle-like morphological characteristics.The mobility of needle-like chrysotile coated by carbonates might be decreased, which probably mitigated the chrysotile's toxicity by changing its morphology.The carbonates also changed chemical composition of asbestos surface.Aust et al. [35] reported that the asbestos toxicity is strongly related to surface chemical composition of respirable elongated mineral particles (e.g., Fe associated with the fibers).The carbonates coated chrysotile may inhibit the Fe-related reaction between chrysotile and biological molecules [35].These results suggest that the aqueous mineral carbonation of asbestos-containing waste slate may reduce toxicity of asbestos.

Conclusions
Leaching and carbonation tests on asbestos-containing waste slate using ammonium salt (CH3COONH4, NH4NO3, and NH4HSO4) solutions were conducted at room temperature and atmospheric pressure conditions.The Ca and Mg-leaching efficiencies increased with decreasing the solid dosage and increasing the solvent concentration.The CH3COONH4 solution had the highest Ca-leaching efficiency (17%-35%) and the NH4HSO4 solution had the highest Mg-leaching efficiency (7%-24%).The NH4HSO4 solution had the lowest Ca-leaching efficiency, probably due to precipitation of gypsum.
The carbonation efficiency determined using the Ca concentrations of the filtrate obtained before and after the carbonation tests was correlated to the Ca-leaching efficiency.The carbonation efficiency in the CH3COONH4 solution ranged between 14 and 27%.The carbonation efficiency was lower in the NH4HSO4 solution due to its lower Ca-leaching efficiency.However, the carbonation efficiency based on the CaCO3 content of the reacted materials was approximately 10%-17% higher than the carbonation efficiency determined using the Ca concentration of the leachate for both the 1 M CH3COONH4 and the 1 M NH4HSO4 solutions because further Ca 2+ ions were dissolved from the waste slate during the carbonation tests.These results suggest that a maximum CO2 sequestration capacity of ~0.07 kg CO2/kg waste slate was obtained at a solid dosage of 50 g/L using the 1 M CH3COONH4 solution.The cost of CH3COONH4 in this direct aqueous carbonation procedure using 1 M CH3COONH4 solution and a solid dosage of 50 g/L can be estimated to be about US$4000 to sequester one ton of CO2 assuming that the price of CH3COONH4 is US$200/ton, which is too expensive.Thus, further study on reducing the cost of chemical is necessary.
The carbonates were precipitated on the surface of chrysotile, which was contained in the reacted materials obtained after the carbonation tests.However, no carbonates were observed on the surface of chrysotile in the as-received waste slate, even though natural carbonation had been occurred in the slate.Consequently, the aqueous mineral carbonation changed the morphology and the surface composition of the needle-like chrysotile by coating it with carbonates.The results of this study suggest that the direct aqueous mineral carbonation of waste slate can be used to sequester CO2 and to reduce the human body risk of asbestos-containing waste slate.However, the results should be verified by conducting multiple tests because a single test for leaching and carbonation was conducted in this study.In addition, further studies are necessary to investigate the human body risk of the solid particles obtained after the carbonation tests and the potential effectiveness of more appropriate conditions to further mitigate the toxicity of asbestos-containing waste slate.

Figure 3 .
Figure 3. Ca and Mg-leaching efficiencies as a function of the solid dosage with the 1.0 M ammonium solution for waste slate samples.

Figure 4 .
Figure 4. Ca and Mg-leaching efficiencies as a function of solvent concentration at the solid dosage of 20 g/L for waste slate samples.

Figure 6 .
Figure 6.Ca carbonation efficiencies as a function of the solid dosage with the 1.0 M CH3COONH4 and 1.0 M NH4HSO4 solutions for waste slate samples.

Figure 7 .
Figure 7. CaCO3 content of reacted waste slate samples determined using TGA as a function of the solid dosage with the 1.0 M CH3COONH4 and NH4HSO4 solutions.
, respectively.No carbonates were observed on the surface of needle-like chrysotile in the as-received

Figure 9 .
Figure 9. Scanning electron microscope (SEM) images of (a) unreacted waste slate and reacted waste slate samples obtained from carbonation tests using the 1 M CH3COONH4 solution at solid dosages of (b) 20 g/L, (c) 50 g/L, (d) 100 g/L, and (e) 150 g/L.

Figure 10 .
Figure 10.Scanning electron microscope (SEM) images of reacted waste slate samples obtained from carbonation tests using the 1 M NH4HSO4 solution at solid dosages of (a) 20 g/L, (b) 50 g/L, (c) 100 g/L, and (d) 150 g/L.

Table 1 .
Chemical and mineralogical characteristics of as-received waste slate sample.

Table 2 .
Results of chemical analysis for filtrates obtained from leaching and carbonation tests.